This invention relates generally to audio reproducers for recorded material and more particularly to such reproducers as are used by stenographers for transcribing the word message into typewritten or similar form. In such systems it is well known that the record is reproduced in a start-stop fashion normally by manual (foot switch) control to enable the typist to listen to a word grouping and type the words prior to listening to the next word grouping. Use of a manually controlled switch while typing necessitates an intermittent typing and listening sequence which is coordinated both with respect to performing the typing and controlling the playback device such that a high degree of skill is required for high speed transcription and the attention required generally results in fatigue for the operator after the process has been sustained for a period of time.